Application of Machine Learning Approaches to Predict Postnatal Growth Failure in Very Low Birth Weight Infants
1
조회
0
다운로드
0
인용
0
참고문헌
키워드
초록
Purpose: The aims of the study were to develop and evaluate a machine learning model with which to predict postnatal growthfailure (PGF) among very low birth weight (VLBW) infants. Materials and Methods: Of 10425 VLBW infants registered in the Korean Neonatal Network between 2013 and 2017, 7954 infantswere included. PGF was defined as a decrease in Z score >1.28 at discharge, compared to that at birth. Six metrics [area under the re ceiver operating characteristic curve (AUROC), accuracy, precision, sensitivity, specificity, and F1 score] were obtained at five timepoints (at birth, 7 days, 14 days, 28 days after birth, and at discharge). Machine learning models were built using four different tech niques [extreme gradient boosting (XGB), random forest, support vector machine, and convolutional neural network] to compareagainst the conventional multiple logistic regression (MLR) model. Results: The XGB algorithm showed the best performance with all six metrics across the board. When compared with MLR, XGBshowed a significantly higher AUROC (p=0.03) for Day 7, which was the primary performance metric. Using optimal cut-off points,for Day 7, XGB still showed better performances in terms of AUROC (0.74), accuracy (0.68), and F1 score (0.67). AUROC valuesseemed to increase slightly from birth to 7 days after birth with significance, almost reaching a plateau after 7 days after birth. Conclusion: We have shown the possibility of predicting PGF through machine learning algorithms, especially XGB. Such mod els may help neonatologists in the early diagnosis of high-risk infants for PGF for early intervention.
Abstract
Purpose: The aims of the study were to develop and evaluate a machine learning model with which to predict postnatal growthfailure (PGF) among very low birth weight (VLBW) infants. Materials and Methods: Of 10425 VLBW infants registered in the Korean Neonatal Network between 2013 and 2017, 7954 infantswere included. PGF was defined as a decrease in Z score >1.28 at discharge, compared to that at birth. Six metrics [area under the re ceiver operating characteristic curve (AUROC), accuracy, precision, sensitivity, specificity, and F1 score] were obtained at five timepoints (at birth, 7 days, 14 days, 28 days after birth, and at discharge). Machine learning models were built using four different tech niques [extreme gradient boosting (XGB), random forest, support vector machine, and convolutional neural network] to compareagainst the conventional multiple logistic regression (MLR) model. Results: The XGB algorithm showed the best performance with all six metrics across the board. When compared with MLR, XGBshowed a significantly higher AUROC (p=0.03) for Day 7, which was the primary performance metric. Using optimal cut-off points,for Day 7, XGB still showed better performances in terms of AUROC (0.74), accuracy (0.68), and F1 score (0.67). AUROC valuesseemed to increase slightly from birth to 7 days after birth with significance, almost reaching a plateau after 7 days after birth. Conclusion: We have shown the possibility of predicting PGF through machine learning algorithms, especially XGB. Such mod els may help neonatologists in the early diagnosis of high-risk infants for PGF for early intervention.
참고문헌 (34)
- Horbar JD, "Weight growth velocity and postnatal growth failure in infants 501 to 1500 grams : 2000-2013", Pediatrics, 136, pp.e84~e92 (2015)
- Lee SM, "Prediction of postnatal growth failure among very low birth weight infants", Sci Rep, 8, pp.3729~ (2018)
- Griffin IJ, "Postnatal growth failure in very low birthweight infants born between 2005 and 2012", Arch Dis Child Fetal Neonatal Ed, 101, pp.F50~F55 (2016)
- Ziegler EE, "Aggressive nutrition of the very low birthweight infant", Clin Perinatol, 29, pp.225~244 (2002)
- Lima PA, "Variables associated with extra uterine growth restriction in very low birth weight infants", J Pediatr(Rio J), 90, pp.22~27 (2014)
- Alur P, "Sex differences in nutrition, growth, and metabolism in preterm infants", Front Pediatr, 7, pp.22~ (2019)
- Ehrenkranz RA, "Growth in the neonatal intensive care unit influences neurodevelopmental and growth outcomes of extremely low birth weight infants", Pediatrics, 117, pp.1253~1261 (2006)
- Franz AR, "Intrauterine, early neonatal, and postdischarge growth and neurodevelopmental outcome at 5. 4 years in extremely preterm infants after intensive neonatal nutritional support", Pediatrics, 123, pp.e101~e109 (2009)
- MacKay EJ, "Application of machine learning approaches to administrative claims data to predict clinical outcomes in medical and surgical patient populations", PLoS One, 16, pp.e0252585~ (2021)
- Shamout F, "Machine learning for clinical outcome prediction", IEEE Rev Biomed Eng, 14, pp.116~126 (2021)
- Goto T, "Machine learning–based prediction of clinical outcomes for children during emergency department triage", JAMA Netw Open, 2, pp.e186937~ (2019)
- Jobe AH, "Bronchopulmonary dysplasia", Am J Respir Crit Care Med, 163, pp.1723~1729 (2001)
- Walsh MC, "Necrotizing enterocolitis : treatment based on staging criteria", Pediatr Clin North Am, 33, pp.179~201 (1986)
- Fenton TR, "A systematic review and meta-analysis to revise the Fenton growth chart for preterm infants", BMC Pediatr, 13, pp.59~ (2013)
- Lin Z, "Quantification of EUGR as a measure of the quality of nutritional care of premature infants", PLoS One, 10, pp.e0132584~ (2015)
- Choi RY, "Introduction to machine learning, neural networks, and deep learning", Transl Vis Sci Technol, 9, pp.14~ (2020)
- Shin D, "Machine learning-based predictive modeling of postpartum depression", J Clin Med, 9, pp.2899~ (2020)
- Cooke RJ, "Improving growth in preterm infants during initial hospital stay : principles into practice", Arch Dis Child Fetal Neonatal Ed, 101, pp.F366~F370 (2016)
- Pilling EL, "Growth patterns in the growth-retarded premature infant", Best Pract Res Clin Endocrinol Metab, 22, pp.447~462 (2008)
- Monset-Couchard M, "Long term outcome of small versus appropriate size for gestational age co-twins", triplets, 2004, pp.89, F310~F314
- Mateus J, "Fetal growth patterns in pregnancy-associated hypertensive disorders : NICHD Fetal Growth Studies", Am J Obstet Gynecol, 221, pp.635.e1~635.e16 (2019)
- Shane AL, "Neonatal sepsis", Lancet, 390, pp.1770~1780 (2017)
- Clark RH, "Extrauterine growth restriction remains a serious problem in prematurely born neonates", Pediatrics, 111, pp.986~990 (2003)
- Han JH, "The impact of neonatal morbidities on child growth and developmental outcomes in very low birth weight infants : a nationwide cohort study", Eur J Pediatr, 181, pp.197~205 (2022)
- Hansson L, "Fluid restriction negatively affects energy intake and growth in very low birthweight infants with haemodynamically significant patent ductus arteriosus", Acta Paediatr, 108, pp.1985~1992 (2019)
- Pavel AM, "A machine-learning algorithm for neonatal seizure recognition : a multicentre, randomised, controlled trial", Lancet Child Adolesc Health, 4, pp.740~749 (2020)
- Gensure RH, "Artificial intelligence for retinopathy of prematurity", Curr Opin Ophthalmol, 31, pp.312~317 (2020)
- Ramgopal S, "Machine learning to predict serious bacterial infections in young febrile infants", Pediatrics, 146, pp.e20194096~ (2020)
- Turova V, "Machine learning models for identifying preterm infants at risk of cerebral hemorrhage", PLoS One, 15, pp.e0227419~ (2020)
- Abbasi H, "Applications of advanced signal processing and machine learning in the neonatal hypoxic-ischemic electroencephalogram", Neural Regen Res, 15, pp.222~231 (2020)
- Slattery SM, "Machine learning mortality classification in clinical documentation with increased accuracy in visual-based analyses", Acta Paediatr, 109, pp.1346~1353 (2020)
- Mboya IB, "Prediction of perinatal death using machine learning models : a birth registry-based cohort study in northern Tanzania", BMJ Open, 10, pp.e040132~ (2020)
- Van Laere D, "Machine learning to support hemodynamic intervention in the neonatal intensive care unit", Clin Perinatol, 47, pp.435~448 (2020)
- Kwon YS, "Development and validation of a quick sepsisrelated organ failure assessment-based machine-learning model for mortality prediction in patients with suspected infection in the emergency department", J Clin Med, 9, pp.875~ (2020)